Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

TCK: expects mutable config sources #304

Closed
tomas-langer opened this issue Jan 10, 2018 · 3 comments
Closed

TCK: expects mutable config sources #304

tomas-langer opened this issue Jan 10, 2018 · 3 comments
Labels
bug 🪲 An error in the implementation code or documentation

Comments

@tomas-langer
Copy link

There are TCK tests that expect system properties to be mutable in MP config. The specification does not state this is a requirement.
ConfigProviderTest.testDynamicValueInPropertyConfigSource()
ConfigProviderTest.testGetPropertyNames()
ConfigProviderTest.testPropertyConfigSource() (as a side effect of properties set during runtime of TCK)
EmptyValuesTest.testEmptyStringProgrammaticLookup()

  1. there should be no requirement for the system properties config source to be mutable
  2. in general mutability of config is dangerous (e.g. if somebody changes 2 values and an application uses both of them, it could get first value from old version and second value from new version of config -> this creates inconsistencies that are really hard to troubleshoot)
@kenfinnigan
Copy link
Contributor

Any update to this?

@dmlloyd
Copy link
Contributor

dmlloyd commented Mar 4, 2020

This relates to #432.

@dmlloyd dmlloyd added bug 🪲 An error in the implementation code or documentation clarification 🏆 A clarification (PR) or request for clarification (issue) in the spec or API and removed mutable clarification 🏆 A clarification (PR) or request for clarification (issue) in the spec or API labels Mar 4, 2020
@tomas-langer
Copy link
Author

As far as I understand the updated specification, it now mandates that Config implementation never caches and must cycle through all config sources to obtain a value.

Although this seems like a major decision about an implementation and a risk (as I mentioned above - if a change happens, we may get data from two change sets), I think this issue can be closed.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
bug 🪲 An error in the implementation code or documentation
Projects
None yet
Development

No branches or pull requests

5 participants